Beyond CTRL F(ind): exploring insights hidden in abstracts through no-code text mining using the example of social entrepreneurship

As the amount of scientific literature continues to grow, it becomes increasingly difficult for scientists to stay up to date within their own domains. Text mining has been suggested as a solution to this problem but has hardly established itself as an alternative to traditional literature reviews in management studies due to the requirement for coding knowledge and familiarity with algorithms. To address this issue, the authors of this paper introduce a no-code solution to text mining based on hierarchical clustering and cosine distance for the clustering of scientific abstracts and titles to create a fine-granular thematic clustering of a research field of choice. To demonstrate the approach, the authors applied it to 2,386 social entrepreneurship abstracts and titles, clustering them into 346 thematic clusters and further categorizing them into 17 different groups. These groups reflect different focus points of the literature, such as sustainability or diversity and inclusion. The authors believe that this approach is valuable both for early-stage researchers, as well as for experienced researchers, as it saves resources and is user-friendly, helping them tackle the ever-increasing amount of literature.